Mike Lantry, 1972
On the heals of another dramatic wins, I dug deep to try and quantify the most dire of Michigan situations that turned into wins from the last ten years. Some are big comebacks, a few are last second heroics. The WPA model is perfect in very late game situations, but it isn’t terrible so here are the ten games from ten seasons that Michigan won based on their lowest WPA during the game for any offensive play.
10. 2004 vs Minnesota, 21% chance of winning
Trailing by 4 with 3 minutes left at their own 13 yard line, Chad Henne led an 87 yard drive to take the lead and the defense held for the win.
9. 2011 at Northwestern, 21% chance of winning
Michigan started the second half trailing by 10 and faced an early 3rd and 11 from their own 19 before Denard hit Roundtree for a first down and Michigan went on to score the next 28 points and win comfortably.
8. 2005 vs Penn St, 17% chance of winning
Michigan took over at their own 40 trailing Penn St 18-10 with under 12 minutes remaining. A 60 yard TD drive with two point conversion and a subsequent Garrett Rivas 47 yard field goal gave Michigan the lead but it came down to that final second when Henne hit Manningham for 10 yards and the winning points.
7. 2010 vs Illinois, 11% chance of winning
In what would become the epic 67-65 swan song victory for Rich Rodriguez, Michigan trailed by a touchdown with five minutes left and had backup Tate Forcier in for an injured Denard Robinson. Forcier led Michigan to a touchdown with 1:47 left and the defense miraculously held. The defense got a final stop on a two point conversion in the third overtime.
6. 2012 vs Northwestern, 9% chance of winning
The punt return from Jeremy Gallon pushed the odds from 5% to nearly 9% but 53 yards later Gibbons trotted onto the field for a chip shot field goal and Michigan pulled it out for the win.
5. 2007 at Michigan St, 8% chance of winning
After a Spartan touchdown, Michigan took over at their own 21 facing a 10 point deficit with seven and a half minutes to go. A Greg Mathews touchdown would cut the margin to 3 in less than a minute. Michigan would force a punt and have a chance to take the lead from their own 35 with four and half minutes left. On 3rd and 12 Chad Henne hit Mario Manningham for a 31 yard touchdown to provide the final margin.
4. 2008 vs Wisconsin, 6% chance of winning
A lone bright spot in a dismal year. The low point came in third quarter with Michigan trailing 19-0. Kevin Grady would convert a 4th and 1 and Steven Threet would eventually hit Kevin Koger for 26 yards to put Michigan on the board. The odds would go back down to 7% before Brandon Minor ran for a 34 yard score, to cut it to a one possession game and bring Michigan’s odds up to 23%. John Thompson’s 25 yard interception return would give Michigan the lead they would hold the rest of the game.
3. 2011 vs Notre Dame, 5% chance of winning
5% is probably high for this one. This is a situation where the late game situations cause the calculator a bit of trouble. No matter what the actual odds where the situation was highly dire and only a Jeremy Gallon invisibility cloak allowed Michigan to pull this one out.
2. 2003 at Minnesota, 2% chance of winning
Michigan entered the fourth quarter trailing Minnesota 28-7. Before scoring 31 fourth quarter points to pull out the win with Garrett Rivas hitting a 33 yard field goal with 47 seconds left to give Michigan its first lead of the game.
1. 2004 vs Michigan St, 1% chance of winning
Down 17 with 8 minutes left at your own 11 yard line. It’s Braylon Edwards time.
Bonus Overtime Chart
+42%-Gardner to Roundtree for 53 yards
+18%-Gardner to Roundtree for 17 yards in OT
+18%-Demens stonewalls Jones to end the game
+18%-Gardner to Funchess for an 8 yard TD
+11%-Gardner to Toussaint for a 28 yard TD
Interesting that none of the top 3 plays were touchdowns
-24%-Gardner intercepted by Dugar
-16%-Pat Fitzgerald come on down, your quarterback is the next contestant on questionable roughing the passer calls
-12%-Siemian to Fields for 21 yards on 1st and 20
-11%-Siemian to Dickerson right before half
-11%-Siemian to Jones to give Northwestern the lead 31-28
Michigan Rush Offense: +5 EV, +26% WPA
Michigan Pass Offense: +11, +64%
Michigan Rush Defense: –6, –9%
Michigan Pass Defense: –8, –44%
Special Teams: +10%
Devin Gardner: +22, +104%
Fitzgerald Toussaint: +1, +13%
Jeremy Gallon: +5, +10%
Roy Roundtree: +7, +65%
Kain Colter: +3, +22%
Trevor Siemian: +9, +41%
Venric Mark: +1, –1%
Dumb Punt of the Week
I really wanted to give this week’s award to our favorite big guy, Charlie Weis for his second half punt from the Texas Tech 38 while trailing but there was another big guy much more deserving. Terry Bowden and the Akron Zips trailed winless UMass by 8 in the fourth quarter and had a 4th and 6 at the UMass 33 yard line. Now a 4th and 6 is not an easy conversion. BUT YOU ARE AT THE 33 YARD LINE! And trailing by 1.5 possessions, in the fourth quarter. The Zips managed to pin UMass at their own 7, but still ended up losing by 8.
Terry Bowden is your Ron Zook Memorial Dumb Punter of the Week
Iowa is not good. Their offense has had one game with an EV+ above 2 (Minnesota) on the season. Their defense has been decent but their three worst performances have come over the last month. No matter who goes at quarterback for Michigan there should be enough firepower to outscore the Hawkeyes. By policy the pick comes from the season numbers but I would consider this score to be a bare minimum margin for Saturday.
Michigan 24 Iowa 10
Let’s head straight to a revamped chart. Now fixed to time, as opposed to play, to give a better feel for the flow of the game.
What jumped out at me right away was how this game was played between 25% and 75% virtually the whole way. In fact, the first play run with either team have a 75% or greater win likelihood was Denard’s completion to The Threat. I combed through my database and Saturday’s game was the longest a game had stayed within that range in the last ten years. No other game had gone 59.5 minutes with neither team being closer to winning than being even. Of course as soon as Michigan’s odds dipped on Toussaint’s ill-advised reception, the offense comes through with a huge completion to set up a 65% chance of hitting the game winning field goal.
Biggest swing plays
Michigan would have been looking at about 70% win odds, but the 26 yards and a new set of downs on Sparty’s fake punt brought Michigan St back to square at 50%.
Andrew Maxwell had a third down and four on Michigan’s side of the field when he threw the ball straight to Jordan Kovacs. Prior to the snap Michigan was at its current low for the game around 39% but the pick and return quickly pushed them to about 53%. The number would have been about 5% higher if part of the return hadn’t been called back.
Michigan was down 1 with the ball at their own 25 with about 5 minutes left. Denard found some room and went 44 yards for Michigan’s longest play of the day. That jumped the game from 44% to 67% in Michigan’s favor.
With less than a minute to go Denard couldn’t find anyone open downfield so he chose to dump it off to Fitzgerald Toussaint a yard behind the line of scrimmage. The ball was low and Toussaint instinctively went down to catch it, which he unfortunately did. The loss of a down, yardage and time pushed the win percent down from 32% to 15%, the first time all game either side crossed the 75% mark.
Michigan would bounce right back and Denard’s strike to The Threat would reverse that 15% in no time. With only a field goal attempt left, the offense handed the game to Gibbons with a kick an average kicker would make 65% of the time.
And of course he did. +35% to Gibbons and all the brunette girls.
[Hit THE JUMP for an updated season projection, Dumb Punt of the Week, Nebraska prediction, and more.]
Back in August anyone who looked at September’s schedule knew a trip to Dallas against the defending national champions and a trip to South Bend made a 2-2 a very real possibility. Here we are in that place and the realist in August doesn’t feel nearly as good about this 2-2 team as he thought he would. 6 turnover days and another Denorges disaster of a game leaves a bad feeling for a lot of Michigan fans.
But maybe it shouldn’t. Alabama appears to have moved into full-on Borg mode, replacing missing pieces and getting stronger together no matter what happens. Michigan’s game may strangely end up being one of the better performances of the season. Notre Dame we all know what happened, but hey, seven point games on the road against quality opponents, especially while going through so much self-inflicted adversity, that’s something that can be built on. And the B1G sucks.
How do all the factors come together, here a look at where the numbers indicate each team versus where they were projected coming in:
Purdue (++): Gone from also-ran to potential spoiler in the Suspendeds division
Minnesota (++): Far from a great team, but a huge step forward from the Gopherquest days
Indiana (+): Still pretty bad but should be more competitive
Michigan St (+): That’s a slight plus from my projection which is probably a double minus compared to most others
Nebraska(0): About where I projected them to be
Ohio St (0): Shaky at times, but still looks like the best in a down conference
Northwestern (0): The wins are great and there will be more, but still not a major player
Iowa (0): A pretty mediocre football team
Michigan (-): Down some, but not a significant variance
Wisconsin (-): Performance has been sloppy to say the least, but the record still holds up
Penn St (-): Have looked a little better the last two weeks but still far from a great team
Illinois: (- -): Getting beat down by Louisiana Tech is not good for the rep
If Ohio wasn’t sitting in timeout and Michigan had gone the cupcake route, no one would be talking about the B1G’s troubles right now. It doesn’t mean they aren’t real but it does skew them. Ohio is the clear-cut favorite to have the most wins in B1G play. They are the only team projected to have more than Michigan’s 5.8 conference wins, with 7.0. Nebraska and Wisconsin are both within a half of the Wolverines while Sparty sits a full game and a half back. Purdue is within a half of game on Wisconsin then it’s a whole pile of garbage. Doesn’t mean a Northwestern or someone couldn’t make a run, but based on current performances, your most likely B1G championship game is Michigan vs Wisconsin.
With high win-probability versus Illinois (98%), Minnesota (84%), Northwestern (96%) and Iowa (92%) still on the board, Michigan’s success is going to come to flipping some of the close ones. Purdue sits at 65%, a low take on the Spartans has that game at 80% (although I can’t imagine anyone has that confidence right now). Nebraska is at 45% and Ohio finishes the year at 20%. That leaves a 3.5% chance of running the table. Beat Nebraska and Ohio and the odds jump tenfold.
Despite two massively frustrating games, with Ohio out of contention, I would still install Michigan as both Legends division and B1G favorites. And this based on in-season performance! If Michigan is able to right the ship, their chances only go up and they are still the most dangerous team eligible for the championship game. With that said, there’s probably another two losses out there for this team. Ohio and Nebraska will be tough road matchups and Michigan isn’t good enough to show up and win the other games on the schedule. A win next week against Purdue would serve as a great firewall for the season.
On to the carnage:
The weekly game probabilities chart
Can you spot the turnovers?
Despite all the big drops in the first half, Michigan was building probability before the critical third quarter fumble.
D. Robinson: –5 EV and –21% Win Probability on 50 plays (was +8 and +12% on 45 of the plays)
F. Toussaint: –1 EV and –6% on 13 plays
E. Golson: –9 EV and –24% on 9 plays
T. Rees: +6 and +14% on 12 plays
Wood/Riddick: +0 and +8% on 24 plays
Could have been nice if Kelly would have stuck with Golson a bit longer. Didn’t have the volume of bad plays of Denard but managed worse before getting pulled for what turned out to be a huge gain in Tommy Rees. As you can see, the running game and Notre Dame offense in general weren’t great, its just the negatives on Michigan’s offense that were the difference.
Over the course of the season there are three key factors that drive the success or failure of a football team:
On a game by game basis you can throw in variance/strategy/luck. It’s pretty tough for a high variance strategy to pay out over the long term but for a particular game playing high variance could be the right decision. Teams like Boise State have found success mostly on execution. Oregon and other non-traditional powers have used offensive systems to drive success. The ones who have done it with talent are easy to spot because that’s where the big-time programs all start.
It would have been unrealistic for Michigan to expect Saturday to be a victory on execution over a Nick Saban coached team. Al Borges was apparently comfortable not pushing any system/variance strategies with his choice of play calling (unless you consider the deep balls his way of playing high variance). That left the major gap between Michigan and Alabama to come down to talent. Michigan and Alabama both have storied histories and bright futures for their football teams, but their current rosters are at very different points.
Here is a look at how Michigan’s roster stacks up to the Big Ten and its non-conference opponents. Methodology here
|Penn St||2,267||Air Force||126|
Michigan certainly has an enviable roster for most of the country, but attrition and recruiting gaps have left the upper-classes of the roster well below the nation’s elite programs. In fact, the gap between Michigan and Alabama is essentially the same as between Michigan and Minnesota. With Michigan not willing (system/variance) or able (execution) to push the other levers, the talent lever came through in full force.
The good news is that there isn’t a team left on the schedule that can do that to Michigan on talent alone. Holding serve on talent puts Michigan at 9-3 and Legends division champs and potentially favored in the B1G Title Game. The talent gap can give and it can take away. Obviously talent is never a guarantee (ask Texas) but with good coaching, Michigan’s talent should put them in a position to be a competitive or win every game remaining on the schedule. The defense seems positioned to possibly pick up some advantage from coaching, and until the full tenants of the passing attack are in place, the offense will likely be middle of the course to slightly above, depending on how the Denard is deployed.
Barring major attrition issues, Michigan will start to move up the talent list over the next several years. I project them to reach current Ohio/ND range in time for the 2014 and potentially hitting the upper echelons when the current freshman enter their senior season. Until that happens, Michigan will either need to be content to see results like they did on Saturday or find different ways to gain advantages over the next two seasons.
Game scores (1st half only)
Denard Robinson: 18 plays, -1.2 EV (points added), -8% WPA (win pct added)
Thomas Rawls: 4 plays, -0.3, -1%
Vincent Smith: 9 plays, -2.0, -1.5%
AJ McCarron: 16 plays, +6.7, +12%
Eddie Lacy: 8 plays, +3.4, +4%
TJ Yeldon: 5 plays, +4.4, +7%
Air Force isn’t charted from its week 1 win against an FCS team, but it plenty efficient, scoring TD’s on 4 of 6 first half drives, with one killed by a 15 yard penalty. The defense pitched a first half shutout but did allow 3 of 5 drives into Falcon territory. My preseason rankings installed Michigan as a nearly three touchdown favorite and I have no reason to think expectations have changed substantially.
Michigan 31 Air Force 10, 98% chance of victory
With the talk of a potentially impending commitment from Laquon Treadwell, Tremendous noted that he might be the most highly touted receiver recruit in Michigan history (extending only to the modern recruiting era, of course).
The prompted me to look at the actual recruiting history of Michigan and fill out a Hall of
Fame Highly Touted roster, ie the most highly regarded players to sign Letters of Intent to Michigan from 2002-2012. Players are ranked only on their recruiting stature. If a player was ranked at different positions by multiple sites, I tried to go with the site that ranked them highest. Without further ado, your Michigan 5-Star Recruiting Hall of Highly Touted.
Michigan’s newest members of the Hall of Highly Touted
Quarterback - Ryan Mallet-(90 points)
Was ranked behind Jimmy Clausen as the number 2 QB in the 2007 class and top 20 overall on three sites. Saw action for an injured Chad Henne during his true freshman season before transferring to Arkansas after Lloyd Carr retired. Was a third round draft choice of the Patriots.
Running Back - Kevin Grady (80)
A consensus five star and top 5 RB in 2005. Showed signs his freshman year after becoming Michigan’s first early enrollee but injuries, legal trouble and the presence of Mike Hart all restricted him from matching his on field results to his profile.
Wide Receivers - Mario Manningham (71) and Antonio Bass (69)
Manningham was Michigan’s most productive wide receiver to come after the 2002 class, twice earning all Big Ten and was a 2nd Team All American in 2007. Was drafted in the third round and made a crucial catch in the Giants Super Bowl win last February.
Antonio Bass saw limited action as a true freshman in 2005 before blowing out his knee in the spring of 2006. Was never able to play again.
Tight End - Will Paul (68)
Scout considered him the #2 Tight End in 2003. Played defensive tackle and fullback at Michigan.
Other skill position - Darryl Stonum (67)
Consensus Top 75 player and Top 15 wide receiver in the 2008 class. Set the single season kickoff return yardage record in 2009 and had a decent 2010 season before a fourth alcohol related incident ended his Michigan career last year.
Offensive Line - Stephen Schilling (73), Justin Boren (72), Kyle Kalis (69), Brett Gallimore (61), Dann O’Neill (58)
Schilling was a sixth round pick in the NFL draft. Boren took his plow and his family values to Ohio. Kalis pulled a reverse Boren and abandoned the Buckeyes and will be a freshman this season. Gallimore switched to defensive line without making much of an impact on either side. O’Neill transferred to Western Michigan after a redshirt season.
Defensive End - LaMarr Woodley (85) and Tim Jamison (69)
Woodley finished his Michigan career with a consensus first team All-American season in 2006 followed by a second round pick in the NFL draft. He was Scout’s #1 ranked defensive end in 2003 and Rivals’ #3 inside linebacker.
Tim Jamison was a top eight defensive end to both services in 2004 but had an under-the-radar career at Michigan. Despite not winning any major postseason awards and going undrafted, Jamison posted the two highest EV season for a Michigan defensive lineman not named Brandon Graham, and has spent three years in the NFL with the Houston Texans.
Defensive Tackle - Marques Slocum (78) and Ondre Pipkins (69)
Marques Slocum only spent a season in Ann Arbor after earning a five star rating from Scout in 2005 but his internet legend will live on forever.
Pipkins comes to Michigan with high expectations of replacing Mike Martin as a true freshman. Hopefully his football career is as great as Slocum’s answers to the quiz.
Linebacker - Brandon Graham (83), Jim Presley (69) and Joe Bolden (61)
Brandon Graham may have suffered through some of the worst defenses Michigan has ever fielded but he did his part to live up to recruiting expectations. Despite ending up on the line, Graham was rated by all the services as a five star linebacker in 2006. Graham was the only player on this list that was ultimately selected in the first round of the draft.
Jim Presley was a four star Top 75 linebacker recruit in 2003 who never was able to cut it academically at Michigan.
Joe Bolden enters his freshman season with Michigan as a consensus four star.
Safety - Prescott Burgess (90) and Jonas Mouton (68)
No Michigan recruit has come with higher ratings than Prescott Burgess. Rivals saw him as the #1 safety in the country in 2003 and Scout saw him as the #3 linebacker. Burgess wasn’t a total bust but never earned more than Honorable Mention All Big Ten and was selected in the sixth round of the NFL draft.
Like Burgess, Mouton ultimately saw his career at Michigan come at the linebacker position. Mouton was up and down at Michigan but was selected in the second round of the draft.
Cornerback - Donovan Warren (74) and Boubacar Cissoko (59)
Warren had three solid seasons before going undrafted as a surprise early entry. Coming out of high school, Warren was considered one of top five corner backs to both Scout and Rivals.
Cissoko paired with Warren to start the 2009 season before legal issues saw him kicked off of the team. His legacy lives on as Teric Jones, Delonte Hollowell and Terry Richardson all committed to come from Cass Tech to Michigan, none of which cracked 5’10.
Defensive line was easily the most competitive position with seven players missing the cut that were rated higher than peers who made it other positions. Chad Henne, Gabe Watson, Will Campbell, Craig Roh, Shawn Crable and Devin Gardner were notable names who were higher than other but not high enough at their position.
Of the 22 players on the list, seven went on to get drafted by the NFL, eight failed to finish their career at Michigan, five finished their career at Michigan without being drafted and Kyle Kalis, Joe Bolden and Ondre Pipkins are yet to begin their careers.
Lloyd Carr brought in nearly all of those players, with between 2 and 4 players from each class from 2002-2007. The 2008 hybrid class had three players on it, none of whom ultimately contributed substantially. In his first full class, Brady Hoke has three new additions to the list and Kyle Bosch and Patrick Kugler are all poised to knock Gallimore and O’Neill from the list once they ink their names in February. If Treadwell ultimately signs then he will knock off Antonio Bass and potentially move in front of Manningham depending on where the final ratings land.
Boubacar Cissoko is far and away the lowest rated player on the list (once the 2013 players join). His consensus value is 59 points. After his spot the two 2013 lineman and Joe Bolden are the lowest rated players, in the low 60s. Everyone else on the list is at least 67 points. Overall Most Touted Recruits are Prescott Burgess for the defense and Ryan Mallett for the offense.
Yes. They still. Even though the Michigan Hall of Highly Touted is a mixed bag tilted more towards busts than wins, I am a still a firm believer that recruiting ratings matter. The important distinction is that recruiting is a volume game, not an individual game. There is still a low rate of individuals living up to their recruiting hype, but if your team is deep with talented players, you should do pretty well.
Two years ago I piled onto the idea that coaches don’t maximize on fourth down. Or at least they don’t maximize to scoring the most points. More likely, they minimize the chance of getting fired for making a risky decision.
Within the post I noted that coaches are too conservative and presented an initial treatment to what the decision matrix should look like. I also identified several of the key objections to the “Stop Punting” movement. Over the last two years I have accumulated more data, thought about the topic further and think there is more to add to the conversation. The general conclusion is the same but I have added some new tools that should shed some new light on the subject and bring some power to the people.
A quick review of the main objections:
4th Down stops don’t account for momentum change
I still haven’t found that this has a quantitative impact to anything. Last month I found that teams go for big passing plays more often but that they don’t actually score any more points.
Assumes all offenses and defenses are average
This is a big one. Two years ago I gave this a surface review but will go into more detail below.
Doesn’t account for in-game situations
It’s a valid critique but one that is rarely applicable and easily approximated. With more than 2 possessions left and a competitive game means expected points are directly correlated with win likelihood. A close game late or a blowout will usually point you where you need to go with common sense. In game situations matter, just less than you think they do.
Men Offenses, Defenses and Kickers are Not Created Equal
This is the big adjustment. Rather than looking at the very best and how their expected value differ from average as I did last time, I made the tool dynamic. I can now look at how each situation changes when you have a dominant offense, or if your defense is a sieve. It can also account for kicker quality. A really good or really bad kicker definitely changes the mindset. Announcers generally frame the decision as “you have to take the three points” even though the odds of success are significantly less than 100%. A great kicker will push the decision closer to that thinking for a quarter of the field. Likewise a terrible kicker takes the field goal out of play for a large portion of the situations.
So to see how different strengths and weaknesses affect the outcomes lets start with the updated baseline chart:
Maximum “go” territory is around the Opponent’s 32 yard line where even on 4th and 10 going for it might be the best play. Once you hit the 20 yard line the field goal is the best option unless you only have one yard to go. This stays true until you get inside the 10, then the opportunity for 3 yards or less becomes more optimal.
So what happens when you get a large variance in teams’ offense and defense. A great example would be the infamous Michigan 2010 team. The offense was as great as the defense and kicking game were terrible. Plugging in extreme values for all three dramatically changes the decision matrix.
Michigan 2010 should have gone for it early and often. Field goals should not have been attempted of 40 yards or more and if the yards to go were 6 or less the offense should have always stayed on the field.
Michigan 2011 was a dramatically different team from 2010. The offense was still good, but not as great as 2010. The defense stopped being a tire fire and the kicking game was much more reliable. I adjusted the offense down a bit and the defense up considerably and moving kicking to average.
The end result is a chart that is somewhere between the baseline and the Michigan 2010. Field goals become the best option when the distance to go is greater but aggressiveness when crossing the 40 is still warranted.
Life, Liberty and the Pursuit of Proper 4th Down Strategy
As part of this work, I have built a public Google Spreadsheet that you at home can use to overanalyze your favorite 4th down situation, complete with dynamic skill ranges.
Each situation adjusts the 4th down conversion chances and upside and downside of any decision. It outputs what the best decision would be and how many expected points you would be leaving on the table with the “wrong” choice.
The offensive and defensive ratings go from +5 to -5 with 0 being average. They are a measure of the difference between the offense and the defense. Michigan’s offense should be a +5 against Minnesota but probably a -2 against Alabama. The rating affects both the odds of conversion and the value of the resulting 1st down if successful. Defense is the same in that it is a measure of the difference between your defense and the opponent offense. Adjusting the defensive metric affects the value of a resulting possession change whether by missed FG, punt or missed fourth down conversion.
I hope you enjoy looking at this and I included some special Easter Eggs for the curious.